Functional genes in the mouse and human genome are associated with CpG islands. Current estimates suggest that there may be as many as 4.5 X 10/4 islands in the human genome and slightly fewer in the mouse genome. These CpG islands are associated with both ubiquitously expressed (housekeeping) genes as well as genes with tissue-specific expression and temporal regulation. The method of restriction landmark genomic scanning (RLGS) identifies rare restriction sites in the genome as landmarks by isotopic end-labeling following enzymatic digestion. These end-labeled fragments are separated in an agarose gel following digestion with a second enzyme. The separated double-digest fragments are digested in-gel with a third restriction enzyme and the resulting products are separated in a final dimension in acrylamide. Genetic variation for these landmarks has been identified between inbred strains of laboratory mice and between common laboratory strains and different Mus species. Variation in landmarks can be mapped in the mouse genome using either standard crosses or recombinant inbred strain analyses. Genetic variation can be used to follow parent-specific alleles in other experimental circumstances. For example, methylation of CpG islands has been associated with the phenomenon of genomic imprinting that results in the epigenetic modification of parent-specific alleles for different genes that leads to a monoparental expression. Parent-specific effects are also important in the analysis of genomic instability associated with tumorigenesis. Loss- of-heterozygosity (LOH) is associated with several kinds of tumors. A simple method of scanning the genome for these kinds of events is an important experimental application of genome analysis. In a similar fashion, genetic variation for these landmarks can be used as markers for the genetic analysis of complex or multilocus traits. In this case, mice from segregating crosses showing specific phenotypes can be pooled for analysis to reduce the experimental effort for identifying genetic differences that are common to the affected progeny. Finally, it is possible to use multiple enzyme combinations of RLGS analyses for genomic regions such as congenic strains to identify candidate genes for a chromosomal region. This is a revised application of a previous proposal to genetically map RLGS loci in the mouse genome. Our primary goal in this proposal is to complete the initial stages of importing the RLGS method into our laboratory from Japan and to establish this as an experimental method for analyzing landmarks that are associated with CpG islands in the genome.